Creating a Segment

What is a segment?

A segment is a subset of data that groups users or accounts on the basis of a specific attribute. This allows you to divide your dataset into smaller chunks for more targeted analysis.

In Loops, segment groups can be created based on a number of criteria or characteristics, such as demographic information, product category, customer behavior, and any other property of an entity. Within each segment group is a specific value, which defines the segment of users who possess a particular attribute. For example: “Country” is a segment group, whereas “USA” is a segment.

Note that segments are mutually exclusive and collectively exhaustive (MECE). This means that each user/account must have a single value for a particular segment – no more and no less.

How to create a segment

1. Log into Loops and click “Data Management” in the left-side navigation bar.
2. Click “Segments” in the left-side menu.
3. Click the Create Segment button at the top-right corner of the screen, and a dialog box will appear.
4. Enter a name and description for your segment.
5. Select a “Category” that reflects the nature of the segment being created. It will be used as a filter in analyses. Options include: Localization, Marketing, Monetization, Technical, and User Properties.
6. Select an “Entity type” based on whether the segment should be composed of individual users or accounts.
7. Check the “Verified confounder” checkbox only if you want this segment to be used in causal models (i.e. be treated as an inherent property of the user/account). Please talk to Loops before using this feature, since inappropriate use will lead to incorrect anlayses.
8. Select the type of segment to be created:
   • A property-based segment allows you to segment users/accounts on the basis of a particular property – that is, the values in a specific column (e.g. “Country”).
   • A condition-based segment allows you to segment users on the basis of a particular condition, and then assign it a name. For example: you can create a segment for when the “Source” column is “like %google%”, and then define the segment name as “Google”.

If creating a property-based segment:

9. Select the column (a.k.a. property) the segment should be based on. You can select from columns that exist within the views that have been created for the entity type you selected. Supported column types include string, boolean and numeric (i.e. float, decimal and integer).
10. Click Save.

Note that Loops will automatically create a “Missing” group for this segment. This group will be visible in future analyses, and will include all users who had an empty or null value in the relevant column.

If creating a condition-based segment:

9. Select the column the segment should be based on, and what conditions should be met (within the column) for a user/account to be part of the segment. You may define multiple segments.

10. Name the segment(s).

11. Name the “Otherwise” segment, which will contain all users/accounts who do not meet the condition(s) for the defined segment(s).

12. Click Save.

Once a segment has been successfully created, it will appear under “Segments” on the Data Management page. From there, you’ll be able to edit, duplicate or delete it.